The epidemiology of Hodgkin's disease (HD) is consistent with a viral etiology. The Epstein-Barr virus (EBV) is closely related to HD on the basis of both serologic and molecular evidence, with 30- 50% of the cases' biopsies being EBV-genome positive. Several risk factors have been consistently identified which point to age of infection as an important modifier of risk in parallel with age at diagnosis. Following the paralytic polio paradigm, both very early and late EBV infections may be more severe. How these risk factor, serologic, and molecular data fit together is unknown. We propose to address this issue in a population-based case-control study conducted in the greater Boston area and the state of Connecticut, involving 600 incident cases and 600 population controls. Risk factor data will be obtained by telephone interview; blood specimens will be drawn from the cases, and tissue blocks tested for EBV- genome. These data will be evaluated for consistency with three models of HD pathogenesis: 1. The EBV is solely related to EBV- genome positive HD with EBV-genome negative disease due to non-viral causes; 2. HD is a virally induced malignancy with the EBV responsible for EBV-genome positive disease and another unidentified virus(es) linked to EBV-genome negative disease; 3. The EBV plays a crucial early role in the pathogenesis of essentially all HD cases but the genome is selectively lost in some patients. Both case-case and case-control comparisons will be done. By integrating the risk factor profile in relation to EBV-genome status of a large series of cases, we should be able to distinguish between these alternative hypotheses. This will be complimented by the serologic comparisons between the two sets of case, and in concert with data from Project 2.